Claude Enterprise is the plan most companies reach for once Claude use spreads past a single team. It answers the identity and audit questions IT asks first. Whether it answers the next question - where the data is processed, and who holds the keys - depends on your security review, not on the plan.
This guide covers what Enterprise includes, how it compares with Team and Max, how the bill is built, and the deployment routes that put Claude inside your own AWS or Azure account instead.
Claude Enterprise is Anthropic's organisation-wide Claude plan: SAML SSO, SCIM provisioning, audit logs, role-based admin controls, custom data retention, higher usage limits, and a contractual commitment that business data is not used for model training by default. It runs on Anthropic's infrastructure and is priced by quote. If your review requires prompts, keys and logs to stay in your own cloud account, the route is Claude in your own AWS or Azure tenant, such as Private Claude Cowork by Moterra.
What is Claude Enterprise?
The name covers three different things, and most guides blur them.
The distinction matters in procurement. A line such as "Claude Enterprise is available on Bedrock, so data stays in your VPC" merges two contracts. The Bedrock bill, the IAM policies and the CloudTrail logs belong to your AWS account. The Enterprise seat, its admin console and its audit log belong to Anthropic's. Both are legitimate choices; they answer different review questions.
What Claude Enterprise includes
The features that separate Enterprise from Team fall into five groups.
Every item above runs in Anthropic's cloud. For many companies that is the right answer. For firms where a regulator, a client contract or an internal policy asks for customer-held infrastructure, it is the point where the next route starts.
Claude Enterprise vs Team vs Max
The models are the same across every option. The choice is a governance choice: who holds the identity, the logs and the keys.
| Max | Team | Enterprise | Your own cloud | |
|---|---|---|---|---|
| Built for | Individual power users | Small teams | Organisation-wide | Regulated firms |
| Billing | Per person, published | Per seat, published | Quoted contract | Your cloud invoice, plus deployment |
| Identity | Individual account | Workspace admin | SAML SSO, SCIM, domain capture | Your IdP, via MDM |
| Audit trail | Personal history | Admin console | Audit logs, Compliance API | CloudTrail or Azure Monitor |
| Where data is processed | Anthropic | Anthropic | Anthropic | Your tenant, your region |
| Encryption keys | Anthropic | Anthropic | Anthropic | You, via KMS or Key Vault |
For a side-by-side of the two organisation-wide routes, see Claude Cowork on 3P vs Claude Enterprise.
Claude Enterprise deployment options
There are five ways to put Claude in front of a business. Only one of them is the Enterprise plan.
Routes three and four give you model access. Route five gives staff the Claude app on top of it, and that is where the deployment work sits: MDM configuration, SSO, connector allowlists and organisation plugins. Private Claude Cowork by Moterra delivers that route in your own AWS or Azure tenant. Moterra manages the deployment; it does not host the data.
Security, compliance, and the questions to ask
Anthropic publishes SOC 2 Type II, ISO 27001 and ISO/IEC 42001 documentation through its trust portal. Certifications settle the vendor question. The route question is settled by eight others.
For UK and EU firms under GDPR and DORA, questions one to four usually decide the route. The detail is in Deploying Claude under GDPR and DORA. For US healthcare, Amazon Bedrock is a HIPAA eligible service, so Claude usage in your own AWS account falls under the BAA you already hold with AWS.
How Claude Enterprise pricing works
Anthropic does not publish an Enterprise price. The contract is quoted on seat count, seat types and expected usage.
Team and Max prices are listed on claude.com/pricing and change often enough that any figure copied into a guide ages quickly. Treat third-party price tables as a starting point and confirm with the account team.
In your own cloud, the bill has two lines. Model consumption is billed per token by AWS or Azure on your existing invoice, visible in Cost Explorer or Azure Cost Management and governed by the budgets you already run. The deployment and management is a per-user fee from Moterra, set by tier, plugins and term. Indicative, final quote confirmed in writing.
Rollout, admin, and licence management
Admin set-up for Enterprise happens in Anthropic's console. In your own cloud it happens in the tools your IT team already runs. A typical Moterra rollout:
Licence management follows the directory. Access is granted by identity provider group, so onboarding and offboarding are the same steps IT already runs for every other system. The 90-day plan in detail is in Claude Cowork for business.
Which option fits your organisation
Frequently asked questions
What is Claude Enterprise?
Claude Enterprise is Anthropic's organisation-wide Claude plan. It adds SAML single sign-on, SCIM provisioning, audit logs, role-based admin controls, custom data retention and higher usage limits on top of the Claude apps, under a quoted contract. It runs on Anthropic's infrastructure.
How much does Claude Enterprise cost?
Anthropic does not publish an Enterprise price. It is quoted on seat count, seat types and expected usage. Team and Max prices are published on claude.com/pricing. Running Claude in your own cloud is billed differently: model consumption on your AWS or Azure invoice, plus the deployment.
What are the usage limits on Claude Enterprise?
There is no published table. Enterprise limits sit above Team and are set in the contract, sized to expected volume. For capacity planning, the number comes from the Anthropic account team. In your own cloud, limits are the service quotas on your AWS or Azure account, which you can raise through the provider.
Does Anthropic train on Claude Enterprise data?
No. By default, Anthropic does not use Claude for Work data, which includes Enterprise, to train its models, and the commitment sits in the commercial terms. On Amazon Bedrock, prompts and outputs are not used to train models and are not shared with the model provider.
Is Claude Enterprise available on AWS Bedrock or Google Cloud?
The plan is not. The Claude models are, through Amazon Bedrock, Google Cloud Vertex AI and Microsoft Foundry, billed by the cloud provider. To give staff the Claude app on top of that model access, Claude Desktop and Cowork are configured against your own cloud account, which is the deployment Moterra runs.
When should I choose Claude Enterprise over Team?
When the rollout is organisation-wide and IT needs SSO, SCIM, audit logs and retention controls. Team suits a small group sharing a workspace. If the security review also requires data, keys and logs to sit in your own cloud account, look at a deployment in your own AWS or Azure instead.
Is Claude Enterprise HIPAA compliant?
Anthropic handles Business Associate Agreement requests for its commercial plans during the sales process. In your own AWS account, Amazon Bedrock is a HIPAA eligible service, so Claude usage there falls under the BAA you already hold with AWS.
How long does a Claude deployment take?
An Enterprise workspace can be switched on in days; the SSO and policy work takes longer. A private deployment in your own AWS or Azure tenant through Moterra is typically live in one to two weeks, including SSO, MDM configuration and core connectors.
Sources · anthropic.com/enterprise · claude.com/pricing · trust.anthropic.com · AWS Bedrock data protection · AWS HIPAA eligible services · Checked 27 Sep 2026
Moterra
Official Anthropic partner. Deploys Claude inside companies' own AWS and Azure tenants.